Event
Ein Rechner für lineare Gleichungen
According to Atanasoff’s later recollection, a prototype was completed from September to November 1939; he and Clifford Berry then began work on the larger machine for systems of linear equations. The Computer History Museum also distinguishes the prototype demonstrated in 1939 from the later machine. Prototype and sources
John Atanasoff and Clifford Berry developed a special-purpose electronic machine for systems of linear equations through 1942. The machine was not programmable. Dating Source
Berry later described two rotating memory drums with capacitors intended to hold coefficients for two equations. Their charge was refreshed when read to preserve stored values. Eyewitness account
Atanasoff’s design manuscript describes storing binary intermediate results on cards; it says work remained in August 1940. In a later letter Berry recalled that the basic computing unit was operating when work stopped in mid-1942, but the circuits for reading binary cards were unfinished. A historical account describes the complete system as unreliable. The design description therefore does not establish fully reliable operation. Design and eyewitness account
Sources
Prototyp 1939 und späterer Rechner: Atanasoffs Rückblick und Museumseinordnung.
Berrys späterer Brief zum Kondensatorspeicher: zwei Gleichungen und Auffrischung.
Entwurfsmanuskript und Berrys späterer Brief: Entwicklungsstand und Grenzen.
Quellenhinweis: Abschnitte Atanasoff-Berry Computer und Clifford Berry.
Offen
Ein detailliertes zeitgenössisches Testprotokoll und die genauen Zustände der Kartenleseschaltungen sind weiter zu prüfen. / A detailed contemporary test record and the exact states of the card-reading circuits remain to be checked.